OCZ Vertex 4 SSD firmware 1.4 has arrived, writes speeds improved, drive speed is now even more amazing

Posted by Paul Braren on May 21 2012 (updated on May 2 2014) in
  • Storage
  • I have the OCZ Vertex 4 256GB SSD (VTX4-25SAT3-256G), used as cache for my RAID array, explained here:

    This drive arrived from the factory with firmware 1.3, so I've been following he firmware 1.4 release candidate discussions, seen here for example:

    and discussed on this PC Perspective Podcast #201 I had recently enjoyed.

    Today, I noticed that OCZ's firmware 1.4 has now come out of beta/release candidate status, and is available here:

    Indeed, just as the OCZ release notes state (dated May 17 2012):


    writes are now faster. Also note that reads are boosted at nearly all sizes, and doubled at 0.5KB! The only slower results were the slight drops seen just at the 512KB & 1024KB read sizes.

    Here's how I tested:

    • took the SATA cable from the RAID controller off, and attached the OCZ Vertex to an AHCI enabled Intel 6Gbps port on vZilla's motherboard, booted, ignoring LSI RAID warnings
    • booted Windows 7 x64 SP1, and used Drive Manager to NTFS format this 256GB SSD
    • downloaded and ran OCZ Toolbox Firmware Updater (that includes firmware 1.4) and didn't just reboot, but instead shut down/powered off (this step is important)
    • ran OCZ Toolbox again, to verify 1.4 was in effect
    • ran ATTO Disk Benchmark, knowing that my 1.3 results were very nearly identical to this review, giving me confidence my 1.4 results were also reliable
    • when I'm done benchmarking/testing, I'll attach the SSD back to my RAID5, and reconfigure CacheCade 2.0 (I don't see a need to run Parted Magic)

    Here's the video of the actual firmware update:

    Check out the results below:


    As far as "AS SSD" benchmarks results, I didn't capture 1.3 results, but here's how firmware 1.4 looks:


    As for my next tests, I'll now put the OCZ Vertex back in the role of RAID5 cache. I'm very curious to witness the firmware boost in that array caching scenario, aren't you?